CN102918526A - Method and apparatus for organizing images - Google Patents
Method and apparatus for organizing images Download PDFInfo
- Publication number
- CN102918526A CN102918526A CN2010800672988A CN201080067298A CN102918526A CN 102918526 A CN102918526 A CN 102918526A CN 2010800672988 A CN2010800672988 A CN 2010800672988A CN 201080067298 A CN201080067298 A CN 201080067298A CN 102918526 A CN102918526 A CN 102918526A
- Authority
- CN
- China
- Prior art keywords
- digital photograph
- trooping
- attribute
- group
- cluster
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
Abstract
A method and apparatus are defined for organizing a plurality of digital photos. The method comprises the steps of identifying a group of digital photos, receiving a number defining how many clusters to be formed from the group, receiving profile information to be used for clustering the digital photos into the number of clusters, clustering the group of digital photos according to the profile information, and identifying representative digital photo(s) of the clusters from the clustered digital photos based on the profile information.
Description
Technical field
The present invention relates generally to the processing of digital photograph.
Background technology
The use of social internet site is becoming more and more general.The population example of whole world larger proportion such as Facebook, twitter exchange with the household with friend with websites such as other social internet sites and keep in touch.Many these internet sites are provided at the possibility of uploading a large amount of photos in the different photo library usually.
In addition, being widely used of digital camera makes the user can take very a large amount of pictures, and need not to consider the cost of processing, and this is to use in the past situation about comprising will be with the video camera of a volume film of every photo certain cost flushing the time.Many mobile phones comprise digital camera, allow thus the user of mobile phone to take pictures in any chance.Because some part has the population of larger proportion to have the mobile phone that comprises digital camera in the world at least, so they always have digital camera on hand.
Needn't worry cost of quality, flushing etc. owing to easily taking a picture, therefore take very a large amount of photos.A shortcoming is that at certain time point, the user may want in some way all his/her digital photographs all to be categorized in classification or the storehouse, and disposes ropy photo.Therefore because the photo amount of classifying may be very large, with their classification and select some to preserve and/or select some tasks of uploading to internet site to bother very much and consuming time.
Just lift an example, family takes a vacation a period of time.During vacation, all kinsfolks may take different photos of many different themes etc. by the different time during a day.When this family went back home, these photos will be classified, and some photos will be selected to upload to internet site, or were stored in catalogue on knee/personal computer/CD or other medium or the file to be shown to friend and family.Suppose to have 5 kinsfolks, and this family week outside, namely 7 days.Suppose that all kinsfolks have taken average 20 photos in the every day in that week.When this family returns, will there be 700 photos to put in order so.Possible this family wants to choose about 50 photos and represents vacation, and this family may want they are shown to friend.
In another example, the user of digital camera and/or comprise that the user of the mobile phone of digital camera may take a picture with short interval such as a plurality of photos every day, but may often not classify to his/her photo, may be every other month once.In this case, the number of pictures that classify may be at the bulk deposition between the moment of classifying.
Proposed and used the whole bag of tricks to come the logarithmic code photo to classify.Existing method and technology adopt the mechanism that is used for analyzing these photos and comes according to their content (being what they describe) they are classified.A shortcoming that is associated with existing method and technology is: the user needs the interested photo of artificial selection, and they may be uploaded to internet site subsequently alternatively.In other words, when these photos for example had been classified in groups, this user need to put these in order not on the same group, and selected interested one or more photo.
Summary of the invention
An object of the present invention is to solve the problem that at least some are summarized above.Specifically, the purpose of this invention is to provide be used to making its user can organize with the least possible artificial treatment the method and apparatus of a plurality of digital photographs.These purposes and other purpose can obtain by the method and apparatus according to following independent claims is provided.
According to an aspect, defined the method that in image organizational equipment, is used for organizing a plurality of digital photographs.The method comprises the steps: to receive and/or preserve the digital photograph group of identifying; Receive definition will form how many clusters (cluster) from this group quantity; Reception will be used for these digital photographs are trooped to the profile information of the cluster of this quantity; According to this profile information this digital photograph group of trooping; And the representative digital photograph of from the digital photograph of trooping, identifying these clusters based on this profile information.
This has following advantage: this digital photograph group will be organized in the cluster with the artificial input of minimum, and will present to this user at least one representative photo of a cluster, some clusters or all clusters, and make him/her can be easily and see rapidly the photo that comprises what type in these clusters.
In one example, this profile comprises attribute, and these attributes represent the different qualities of photo.
According to an embodiment, these attributes comprise the characteristic of the outward appearance of the metadata of photo and/or relevant described photo.This has following advantage: based on photo portrayal what or the outward appearance of photo and the certain metadata of digital photograph, these digital photographs can be organized or be categorized in the cluster.
According to another embodiment, the method also comprises: weighted associations is arrived some or all of attributes, thereby define the importance of each attribute.This has following advantage: can define more accurately cluster so that can high precision or pin-point accuracy carry out the tissue of these digital photographs.
In one example, this profile information comprises definition for the user profiles of the effective community set of specific user, and these attributes are defined in the different qualities that will consider when trooping this digital photograph group.This has following advantage: the user definition of giving the method is best corresponding to the ability of the profile of his/her personal preference and preference.
In another example, this profile information comprises definition for the situation profile of the effective community set of particular condition, and described attribute is defined in the different qualities that will consider when trooping this digital photograph group.This has following advantage: the user can will watch these digital photographs or define different profiles according to the particular condition that different digital photographs are described according to whom.
According to an embodiment, the particular community that will consider when the step of the profile information that receives these digital photographs that will be used for trooping is trooped described digital photograph group by being logged in creates this situation profile.This has following advantage: make this process the user can or the typing certain profiles, create new profile, perhaps just typing is wanted more expendable attributes for these digital photographs of trooping.
According to another embodiment, the method also comprises: before the step of the described digital photograph group of trooping according to described profile information, determine the determinacy of different attribute with respect to different digital photographs, and leach the digital photograph that has uncertain information for particular community.
In one example, comprise according to one or more attribute definition clusters according to the troop step of described digital photograph group of described profile information, and according to the described cluster described digital photograph group of trooping.
In another example, the method also comprises: add the digital photograph that leaches, have uncertain information to suitable cluster after the step of the representative digital photograph of identifying described cluster based on described profile information from the digital photograph of trooping.This has following advantage: the digital photograph that leaches is not lost, and adds them after having identified representative photo.It is also guaranteed: the digital photograph that is difficult to troop can not be identified as the representative digital photograph of cluster.
According to an embodiment, determine that with respect to different digital photographs the determinacy of different attribute comprises: determine the entropy of these attributes, wherein high entropy is corresponding to uncertain information.
According to another embodiment, the step of identifying representative digital photograph comprises: identify the photo in each cluster, this photo the best is corresponding to the attribute of defined that cluster in this profile information.
According to another embodiment, if typing user profiles and in the step of the described digital photograph group of trooping according to this profile, use subsequently this user profiles, then the method also comprises: receive from the feedback of the representative digital photograph of the user, relevant identification and adjust the weight of each attribute in the user profiles according to the feedback that receives.This has following advantage: can " repair " or optimize this user profiles to be fit to this user's the expectation and the preference that are associated with this profile.
In one example, the step that receives and/or preserve the digital photograph group of identifying comprises: described digital photograph group is uploaded to data storage device.
In another example, the step that receives and/or preserve the digital photograph group of identifying comprises: at least one digital photograph uploaded to the data storage device that comprises the digital photograph set of clusters that has existed, and wherein saidly comprise according to the described profile described digital photograph group of trooping: at least one digital photograph of uploading is trooped in one of cluster in the digital photograph set of clusters that has existed.This has following advantage: the user can add single digital photograph or several digital photograph, and they are organized in the group of trooping or organizing of digital photograph.
According on the other hand, define a kind of equipment that is suitable for organizing a plurality of digital photographs.
In one embodiment, described equipment comprises: database is suitable for receiving and/or preserving the digital photograph group of identifying; Receiving element is suitable for receiving the input from the user, and described input comprises the number of clusters that will form and will be used for described digital photograph is trooped to the profile information of the cluster of described quantity from described digital photograph group; The unit of trooping is suitable for according to the described profile information described digital photograph group of trooping; And recognition unit (520), be suitable for identifying from the digital photograph of trooping based on described profile information the representative digital photograph of described cluster.
According to an embodiment, described profile information comprises attribute, and described attribute represents the different qualities of digital photograph.
In one example, described attribute comprises the characteristic of the outward appearance of the metadata of photo and/or relevant photo.
In another example, weighted associations is arrived some or all of attributes, thereby define the importance of each attribute.
In another example, the profile information of reception be definition for the user profiles of the effective community set of specific user, described attribute is defined in the described different qualities of trooping and will consider during described digital photograph group.
In another example, the profile information of reception be definition for the situation profile of the effective community set of particular condition, described attribute is defined in the described different qualities of trooping and will consider during described digital photograph group.
According to an embodiment, when receiving described trooping, receiving element creates described situation profile during the particular community that will consider during described digital photograph group.
According to another embodiment, this equipment also is suitable for: described digital photograph group is trooped in the described unit of trooping according to described profile information before, determine the determinacy of different attribute with respect to different digital photographs, and leach the photo that has uncertain information for particular community.
According to another embodiment, comprise according to the described profile information described digital photograph group of trooping: according to one or more attribute definition clusters, and according to the described cluster described digital photograph group of trooping.
In one example, this unit of trooping also is suitable for: add described digital photograph that leach, that have uncertain information to suitable cluster after the representative digital photograph of identifying described cluster based on described profile information from the digital photograph of trooping.
In another example, determine that with respect to different digital photographs the determinacy of different attribute comprises: determine the entropy of these attributes, wherein high entropy is corresponding to uncertain information.
In another example, identify described representative digital photograph and comprise: identify the photo in each cluster, this photo is best corresponding to the attribute of defined that cluster in this profile information.
According to an embodiment, if described receiving element receives user profiles (it is used when trooping described digital photograph group according to described profile by the described unit of trooping subsequently), then this receiving element also is suitable for receiving from the feedback user, the relevant representative digital photograph of identifying, and this unit of trooping also is suitable for adjusting according to the feedback that receives the weight of each attribute in this user profiles.
According to another embodiment, this database is suitable for preserving the group of the digital photograph of having trooped and receives at least one digital photograph, and wherein this unit of trooping also is suitable at least one digital photograph of described reception is trooped in one or more clusters in the group of the digital photograph of having trooped.
According to another embodiment, described equipment is terminal, such as digital camera, mobile phone or comprise any other terminal of video camera.
Description of drawings
To and be described in greater detail with reference to the attached drawings the present invention by example embodiment now, in the accompanying drawing:
Fig. 1 is the process flow diagram of illustrative process embodiment.
Fig. 2 is the process flow diagram of another embodiment of illustrative process.
Fig. 3 is the process flow diagram of the another embodiment of illustrative process.
Fig. 4 is the block diagram of illustration profile.
Fig. 5 is the block diagram of exemplary apparatus embodiment.
Fig. 6 is the entropy curve.
Embodiment
This solution can be used for solving the problem that at least some are above summarized.Specifically, this solution can be used for making the user a plurality of digital photographs can be organized into automatically cluster (cluster) or group.
In the prior art, the user must artificial selection he/her want to be placed on photo in photograph album or the file.He/her also need to be before picture be uploaded to photograph album or file artificial cognition different crucial moment, that is, for example represent the picture of a certain event.
Solution given here will provide helps the user with minimum manual intervention and according to different attribute a plurality of photos is organized into process and equipment in the different clusters automatically.
Fig. 1 is process flow diagram, and this process flow diagram illustration is for the embodiment of the process of organizing a plurality of digital photographs.This process comprises the first step 100 that receives and/or preserve the digital photograph group of identifying.This can carry out with various ways.Two examples are that the group of photos that will organize uploads to database, or identification is included in the group of photos in data storage device or the storer.
The next step 110 of this process is to receive how many clusters definition will form from this digital photograph group quantity.A feature of this process is that it is with a plurality of digital photographs (referred to herein as the digital photograph group) tissue or troop in less group (hereinafter referred to as cluster).The group of photos of organizing may be very large, and it comprises a large amount of digital photographs.In order to carry out this feature of this process, that is, this digital photograph group is organized into cluster, be necessary to know from this digital photograph group, to form how many clusters.
Fig. 1 the has gone back illustration next step 120 of this process, that is, and reception will be used for the trooping profile information of these digital photographs.This step can be known this process how to troop this digital photograph group.What in order to carry out this digital photograph cohort collection in cluster, also be necessary to know how to troop to comprise the digital photograph of type in this group of photos, each cluster.This information will be provided by profile information.
Useful different modes is realized the typing of profile information.In one example, the profile of these digital photographs but these process prompting user 220 typings will be used for trooping.To be explained in more detail in conjunction with Fig. 2 hereinafter.
Then, trooping of this digital photograph group 130 can begin, and according to the profile information of typing in previous step 120 digital photograph is clustered in the different clusters.By this, this process has had it with this digital photograph cohort collection or has been organized into needed all information of cluster.Number of clusters that the group that will troop, this digital photograph group will troop/be organized into, with good grounds which standard or how to carry out this and troop (as given by this profile information) are also provided to this process.
Fig. 1 has gone back illustration and has identified the final step 140 of the representative photo of these clusters based on the profile information of typing from these digital photographs of trooping.When the original photo group was trooped in the different clusters, this step provided representative photo for cluster.Then he/her can provide to the user of the method the photo of the specific cluster of representative, so that will know or easily see that the photo of what classification is comprised in that specific cluster.May there be the one or more photos that represent cluster.All clusters or some of them cluster can be represented by representative photo.
This has a plurality of advantages.An advantage is: this user can be automatically with a large amount of digital photograph classification.Only need to stipulate number of clusters and profile information by this user, and then this digital photograph group or a plurality of digital photograph by automatic classification or tissue.Then will present to this user at least one representative photo of a cluster, some clusters or all clusters, and make him/her can be easily and see rapidly the photo that in these clusters, comprises what type.This user needn't manual sorting and is selected a large amount of photos, and/or artificial selection tissue or the various criterion that will use when trooping these digital photographs.In addition, will provide to this user the representative digital photograph of cluster, check the digital photograph that in these different clusters, comprises what type so that this user needn't manually open each cluster.
By typing 120 and the profile information that is used for subsequently trooping 130 these digital photographs and identifies the representative photo of 140 these clusters can comprise attribute, described attribute represents the different qualities of digital photograph.
Attribute can be can indicate or define digital photograph to describe and so on any category Properties.The most often expect group of photos is made into cluster, wherein cluster can comprise the photo that discloses similar sight and/or object.These attributes help to be defined in the digital photograph that should comprise what classification or type in each cluster.
An advantage of this feature is: the one or more profiles of user creatable, and each profile comprises attribute, this attribute represents the different qualities of photo.Another advantage is: when this profile information of typing, but the profile that user's typing will be used, and need not each attribute of manual entry when he/her wants to organize the digital photograph group.
These attributes comprise the characteristic of the outward appearance of the metadata of photo and/or relevant photo.
May be advantageously: when definition profile and/or in cluster step and identification step subsequently, will use what time, can according to metadata and what (outward appearance of photo) of this photo portrayal both stipulate a plurality of different attributes.Some examples of metadata are time and dates of pictures taken.May expect these digital photographs are sequentially classified according to certain that when is relevant to that they are taken, or by one day time (such as day or night) classification.Other example of some of attribute is what photo shows, such as people, smiling face, landscape, mountain, buildings etc.Some other examples are take a picture local and along which direction.Another attribute can be the acutance of this photo, may expect not show fuzzy photo.Certainly, may exist photo portrayal what with the many different examples that can consider what metadata.
These different attributes also can be associated from different weights, thereby define the importance of each attribute.Weight can be associated with some different attributes or all different attributes.
This also has a plurality of advantages.Some attributes may be more important than other attribute, and should be considered especially in cluster step and identification step.This feature makes the user of this process can be better and the different profiles (and the cluster in these profiles) that will use of more accurately definition.The quality that this also can improve this cluster step and particularly also have identification step so that cluster more likely only comprise should be as the photo of the part of that cluster.In identification step, be identified as the digital photograph of the specific cluster of representative more likely best corresponding to the most important attribute of that specific cluster, the attribute that namely has highest weighting.
These attributes can be expressed as probability, its scope from 0 to 1.To be described in more detail this hereinafter.Can make up different attribute.For example, can composite attribute " smiling face ", " place " sort out with the digital photograph that will describe identical or similar object with " direction ".When organizing/trooping these digital photographs, the combination of these attributes will be served as the basis of choice criteria.
Digital photograph p is by its attribute
Att Xl Representative, so that: p=
Similarity sim (p between two photos
1, p
2) can be expressed as:
Wherein ont is the body of the attribute ratings of similarity measurement between the relevant defined attribute, and w
kAttribute att
kWeight.Similarity between the different attribute can be calculated as follows according to the value of this attribute: " similarity between the number ": the standardization distance between 0 and 1, or equivalent similarity; " similarity between the binary value ": if they are equal, then are 1, otherwise are 0, perhaps equivalent similarity; " similarity between project and the tabulation ": during given operational set (such as insertion, deletion etc.) first module is transformed into second unit or second unit is transformed into the required step number of first module; " similarity between the grading list ": begin similar step number from root.These weights are by standardization, and these weights will reflect the importance of each attribute when trooping these photos.
Can be that definition is for the user profiles of the effective community set of specific user by the profile information of typing 120.These attributes are defined in the different qualities that will consider in the step 130 of this digital photograph group of trooping.
By user's profile, give the following ability of user of this process: definition is best corresponding to the profile of his/her personal preference and preference.The user probably about he/her to what interested and he/her prefer to see what has some preferences.Thereby user profiles can be to be user " that carries out is special ".
Can be that definition is for the situation profile of the effective community set of particular condition by the profile of typing 120.These attributes are defined in the different qualities that will consider in the step of this digital photograph group of trooping.
By having the situation profile, but the different situation profiles of user's typing of this process.May expect according to whom will watch these digital photographs or use different profiles according to the certain scenarios of describing in these digital photographs.May expect dissimilar photo (situation) is shown to friend, and the photo of other type is shown to household and relatives.So, these situation profiles will comprise different attribute.Also may be following situation: some digital photographs can be uploaded to social networks, and some digital photographs will be stored in the particular file folder or database on the storer.Modification has many, and may be basic for the needs with different profiles therefore.
This situation profile can be by pre-defined.But also can create this situation profile in the typing particular community that the step of profile information of these digital photographs will consider in by the step that is logged in this digital photograph group of trooping that will be used for trooping.
Corresponding to " well " profile or the matching profile that will be used for this expectation attribute of trooping, can typing to may be expectation and favourable for some particular communitys of these digital photographs of trooping then if there is no.This make this process the user can or the typing certain profiles, create new profile, perhaps just typing is wanted more expendable attributes for these digital photographs of trooping.Certainly, it also makes the user of this process can create new profile, and these attributes of typing for example can be used as the situation profile and store.
Fig. 2 is the process flow diagram of another embodiment of this process of illustration, wherein realizes this process with prompting user typing profile 220 or particular community as this mode of profile information.If this user can't so do, then troop 260 and identification 270 steps in, predefined user profiles 240 is retrieved and used to this process automatically.Step 200 among Fig. 2 and 210 is corresponding to the step 100 among Fig. 1 and 110.
In an embodiment of this process, such as institute's illustration among Fig. 2, possiblely be: before the step 260 of this digital photograph group of trooping according to the profile of typing, determine the determinacy of different attribute and leach 250 to have the photo of uncertain information for particular community.The step 260 of Fig. 2 is corresponding to the step 130 of Fig. 1.
This can have a plurality of advantages.Expectation is according to the profile information of typing or attribute this digital photograph group of trooping.More possible digital photographs are difficult to classification or troop, and have certain uncertainty with respect to different attribute.These photos should not be identified as the representative photo of cluster, and this is avoided thus.
Attribute in this profile or this profile information can be counted as or be expressed as following probability: photo portrayal what and/or metadata.Some digital photographs will more easily be classified or troop than other photo.As an example, for digital photograph, the probability that it describes the smiling face is 0.5.This means, the probability that it does not describe the smiling face also is 0.5.Therefore, smiling face's attribute is uncertain high.This digital photograph will be filtered off 150.Suppose on the other hand: for digital photograph, the probability that it describes the smiling face is 0.8.This means, the probability that it does not describe the smiling face is 0.2.Smiling face's attribute is uncertain low in this number code photo, and therefore this digital photograph is easy to troop with respect to attribute " smiling face ".
Comprise according to the troop step 130,260 of this digital photograph group of this profile information: according to one or more attribute definition clusters, and according to defined cluster this digital photograph group of trooping.
As mentioned above, this profile information comprises different attribute.This profile information also defines different clusters, and these clusters respectively comprise again one group of one or more attribute.When trooping 130,260 this digital photograph group according to this profile information, according to the cluster that in this profile information, defines this digital photograph group of trooping.
After identification step 270, when identifying the representative photo of these clusters based on the profile information of typing from these digital photographs of trooping, this process comprises that the digital photograph that will leach, have uncertain information adds 280 steps to suitable cluster.
This has following advantage: the digital photograph that leaches is not lost, and can suitably find in the cluster.Suitably cluster is according to this digital photograph of attribute of that cluster the most corresponding cluster.Digital photograph with uncertain information does not mean that it is " bad " photo or undesired photo.It is trooped the uncertain information indication with can not determine.For example, suppose to troop according to attribute be the smiling face.This photo probably comprises the smiling face, but this process can not be determined beyond the question this point or determine this point with high probability.Therefore, this photo is filtered off, but is not dropped.After the identification representative photo, this digital photograph is added to the digital photograph cluster that comprises the smiling face.As a result, this digital photograph still can be found by this user, but it can not be the representative photo of that specific cluster.
The determinacy of determining different attribute with respect to different digital photographs comprises: determine the entropy of these attributes, wherein high entropy is corresponding to uncertain information.
Example in the reference, for a photo, wherein it comprises that smiling face's probability is 0.5.In this example, the entropy of attribute " smiling face " is high.Equally, be 0.8 photo for its probability of comprising the smiling face wherein, the entropy of attribute " smiling face " is low.
Entropy can be used as probabilistic measuring in the sample set.It can be said to be the impurity level in the random collection of characterization example, and it for example can be expressed as stochastic variable Y.Expectation self-information (HY) then can be expressed as:
In this set only 2 values (positive and negative) and suppose we have 14 samples (wherein 9 just negative with 5
) situation under, entropy E becomes:
If this set has equal distribution, for example [5+, 5-], then E (Y)=1.If all samples all belong to same class, for example [10+, 0-], then E (Y)=0.See the chart among Fig. 6, it discloses the entropy curve.
The step 270 of identifying at least one representative digital photograph comprises: identify in each cluster best the photo corresponding to the attribute of defined that cluster in this profile information.
At least one digital photograph that presents each cluster to the user of this process is favourable.Photo is the photo of close barycenter corresponding to the digital photograph of the attribute of defined that cluster in this profile information best, and this barycenter is 100% corresponding to the photo by the defined cluster of its attribute.As explained above, attribute can be counted as existing in the digital photograph probability of that attribute.100% will have 1.0 probability corresponding to the digital photograph by the defined cluster of its attribute for each attribute in that cluster.In other words, the entropy of each attribute will be 0 in that cluster.Having this digital photograph that equals 0 entropy for each attribute in that cluster may not exist, and it can be regarded as " hypothesis " digital photograph.The representative digital photograph of cluster will be the photo of the most close " hypothesis " digital photograph, namely have the digital photograph of minimum entropy for the attribute of this cluster of definition.Can in the step 280 of the representative digital photograph of identification, consider other individual standard, such as the quality of digital photograph, resolution etc.
When the user to this process is the digital photograph of representing cluster, this user also will have the possibility of more photos of seeing that cluster.This user will be not limited to only see representative photo.This process can be embodied as so that this user can be chosen in the digital photograph that the representative digital photograph that has been presented in the specific cluster is presented any amount in the specific cluster afterwards, or by all digital photographs in that cluster of this representativeness digital photograph representative.
This process also comprises following possibility: this user's typing is about trooping and the feedback of the photo of identifying.This illustration in Fig. 3.If receive user profiles and in the cluster step 130,260 of this digital photograph group of trooping according to the user profiles of typing, use subsequently 300 these user profiles, but the feedback 310 of the relevant representative photo of identifying of this user's typing then.Then can adjust according to the feedback that receives the weight of each attribute in 320 these user profiles.
This has following advantage: can " repair " or optimize this user profiles to be fit to user's the expectation and the preference that are associated with this profile.The preference of this user profiles representative of consumer is and if this profile is adjustable is made into so that this process according to most probable user preference this digital photograph group of trooping, then is favourable.When adjusting 320 these weights, then store 330 user profiles that upgrade.
How to upgrade weight w
KlAn example be:
,
Wherein v is current observation.
According to this process, the step 100,200 that receives and/or preserve digital photograph group that identify, that will organize comprises: this digital photograph group is uploaded to data storage device.
One of them is which photo of identification will be organized for the primary thing that need to do when organizing a plurality of photo.According to the realization of the method, as will be described later, this can carry out with various ways.This for example understands a kind of mode of the digital photograph group that identification will be organized and troop.In other words, the group of photos that will organize of the user of this process uploads to certain type data storage device carrying out this process thereon.
According to an embodiment of this process, the step 100,200 that receives and/or preserve digital photograph group that identify, that will organize comprises: at least one digital photograph uploaded to the data storage device that comprises the cluster of photographs group that has existed.Then, comprise according to the troop step 130,260 of this digital photograph group of this profile information: at least one digital photograph of uploading is trooped in one of cluster in the digital photograph set of clusters that has existed.
This has consisted of how to receive and/or preserve the digital photograph group 100 of identifying, will organize, another example of 200.In this example, this group of photos that troop comprises that the digital photograph set of clusters that has existed adds the digital photograph that at least one is uploaded.Then, in this cluster step 130,260, this process only must be trooped newly uploaded at least one digital photograph in the digital photograph set of clusters that has existed.In other words, what this can be counted as increasing progressively troops, and wherein the user can add digital photograph to the digital photograph set of clusters that has existed simply.An advantage of this example is: the digital photograph set of clusters that all have existed does not need again to troop.Only new at least one digital photograph that adds needs processed in order to it is trooped to suitable cluster.Certainly, this only uses same profile information just effective in the situation that should troop for this.If other profile information of the user selection of this process is used in the cluster step 130,260, then needs are considered all these digital photographs when trooping according to the profile information of new typing.
Fig. 4 illustration the example of user profiles 400 or situation profile 400.This profile comprises different attribute
Respective weights with each attribute
This means, can give its importance more or less by apply weight to each attribute.In other words, each attribute has A
k* W
kImpact.
The equipment that is suitable for organizing a plurality of digital photographs is described below with reference to Fig. 5.This equipment has advantages of similar with the method or identical, and these will not repeat.Fig. 5 discloses the block diagram of exemplary apparatus embodiment.
The equipment 500 of illustration comprises the database 510,511 that is suitable for receiving and/or preserving the digital photograph group of identifying among Fig. 5.This database or can be the part (such as 510 illustrations of database) of equipment 500, perhaps it can be embodied as external data base 511.Equipment 500 also comprises the receiving element 540 that is suitable for receiving from user's input, and this input comprises the quantity of the cluster that will form and will be used for digital photograph is trooped to the profile information of this quantity cluster from this digital photograph group.Equipment 500 also comprises and being suitable for according to the troop unit 530 of trooping of this digital photograph group of this profile information, and it also comprises the recognition unit 520 that is suitable for coming based on this profile information the representative digital photograph of these clusters of identification from the digital photograph of trooping.
Receiving element 540 comprises attribute from the profile information of user's reception of equipment 500, and described attribute represents the different qualities of digital photograph.
This profile information can be stored in this equipment, and this user only typing to the reference of the profile information that will use.When equipment 500 received the reference of this profile, it can retrieve institute's stores profile information in order to use.Alternatively, this profile information can be by user's manual entry of equipment 500.
The attribute of this profile information comprises: the metadata of photo and/or the characteristic that derives from analyze digital photograph.
Some examples of the characteristic that derives from analyze digital photograph are smiling face, people, landscape, buildings, mountain, seabeach, animal etc.
The some or all of attributes of the profile information that is received by receiving element 540 in one embodiment, are associated with the weight that defines each Importance of Attributes.
In one example, the profile information that is received from the user of equipment 500 by receiving element 540 be definition for the user profiles of the effective community set of specific user, these attributes are defined in the described different qualities of trooping and will consider during this digital photograph group.
In another example, the profile information that is received from the user of equipment 500 by receiving element 540 is that definition is for the situation profile of the effective community set of particular condition, the different qualities that will consider when these attributes are defined in this digital photograph group of described cluster.
In one example, troop during the particular community that will consider during this digital photograph group when receiving element 540 receives described, create the profile that is received from the user of equipment 500 by receiving element 540.
Then equipment 500 be suitable for creating profile and storing the profile that creates according to the attribute that receives.
In one embodiment, the unit 530 of trooping of equipment 500 also is suitable for: this digital photograph group is trooped in the unit 530 of trooping according to this profile before, determine the determinacy of different attribute with respect to different digital photographs, and leach the photo that has uncertain information for particular community.
In one example, when this digital photograph group is trooped according to profile information in the unit 530 of trooping, define cluster according to one or more attributes, and the unit 530 of trooping is according to these clusters this digital photograph group of trooping.In other words, the unit 530 of trooping is suitable for: when trooping this digital photograph group according to this profile information, and according to one or more attribute definition clusters, and according to these clusters this digital photograph group of trooping.
In one embodiment, the unit 530 of trooping of equipment 500 also is suitable for: add the digital photograph with uncertain information that leaches to suitable cluster after the representative digital photograph of identifying these clusters based on this profile information from the digital photograph of trooping.
In one embodiment, wherein the determinacy of different attribute is determined in the unit 530 of trooping of equipment 500 with respect to different digital photographs, and this comprises: determine the entropy of these attributes, wherein high entropy is corresponding to uncertain information.
The representative digital photograph of recognition unit 520 identification clusters.This identification comprises: identify in each cluster best the photo corresponding to the attribute of defined that cluster in this profile information.
If receiving element 540 receives user profiles (it is used when trooping this digital photograph group according to this profile by the unit 530 of trooping subsequently), then receiving element 540 also is suitable in one embodiment: receive from the feedback user, the relevant digital photograph of identifying, and the unit 530 of trooping also is suitable for: the weight of adjusting each attribute in this user profiles according to the feedback that receives.
In one embodiment, the database 510,511 of equipment 500 also is suitable for: preserve the group of the digital photograph of trooping that has existed and receive at least one digital photograph.Then the unit 530 of trooping also can be suitable for: the digital photograph of at least one reception is trooped in the one or more suitable cluster in the group of the described digital photograph of trooping that has existed.
Advantageously, equipment 500 can be any terminal that comprises video camera.In this case, will in this terminal, realize aforesaid method.The terminal that comprises video camera generally includes data storage device (such as storer), to preserve the digital photograph of being taken by its video camera.Then the user of this terminal can utilize this process after having taken a plurality of photos, and according to certain profiles or according to particular community at these these photos of terminal inner tissue.This user also can incrementally be organized into the digital photograph of newly clapping in the suitable cluster that has existed.
If the user of (all as described above) terminal wishes digital photograph is uploaded to social internet site, such as for example Facebook, then he/her can easily select to upload the representative digital photograph of whole cluster, part cluster or that cluster.
This equipment is realized in also can the node in network providing access or use to the method with the user to social internet site.Then the user can upload to social internet site with a plurality of photos, and organizes these photos according to the method in this internet site.Then, can give which photo of this user selection and deliver and make other people appreciable option.Also have, in this example, this internet site may have been preserved a plurality of digital photographs of having trooped, and this user can upload to this internet site with one or more digital photographs, and is organized in the cluster that has existed it or they.
In another example, this equipment can be realized in personal computer or laptop computer.As for above-described other example, the user can upload a plurality of photos and organize them in this laptop computer.As in other example, this computing machine or laptop computer may have been preserved a plurality of digital photographs of being trooped, and this user can upload to this computing machine or laptop computer with one or more digital photographs, and is organized in the cluster that has existed it or they.
Should be noted that Fig. 5 only on logical meaning illustration the various functional units in this equipment.Yet those skilled in the art can use any suitable software and hardware member freely to realize these functions in practice.Thereby, the present invention generally be not limited to this equipment and these functional units shown in structure.
Although described the present invention with reference to particular exemplary embodiment, instructions is generally only planned the illustration inventive concept, should not be regarded as limiting the scope of the invention.The present invention is defined by appended claims.
Claims (30)
1. method that is used for organizing a plurality of digital photographs in image organizational equipment, described method comprises the steps:
The digital photograph group that-reception and/or preservation (100,200) are identified,
-receive (110,210) to define the quantity that will from described group, form how many clusters,
-receive (120,230,240) and will be used for described digital photograph is trooped to the profile information of the cluster of described quantity,
-according to described profile information (130,260) the described digital photograph group of trooping, and
-from the digital photograph of trooping, identify the representative digital photograph of (140,270) described cluster based on described profile information.
2. the method for claim 1, wherein said profile comprises attribute, described attribute represents the different qualities of photo.
3. method as claimed in claim 2, wherein said attribute comprise the characteristic of the outward appearance of the metadata of photo and/or relevant described photo.
4. such as each described method among the claim 1-3, also comprise: weighted associations is arrived some attributes or all properties, thereby define the importance of each attribute.
5. such as each described method among the claim 1-4, wherein said profile information comprises definition for the user profiles of the effective community set of specific user, and described attribute is defined in the described different qualities of trooping and will consider during described digital photograph group.
6. such as each described method among the claim 1-4, wherein said profile information comprises definition for the situation profile of the effective community set of particular condition, and described attribute is defined in the described different qualities of trooping and will consider during described digital photograph group.
7. method as claimed in claim 6, wherein receiving (120,230,340) to be used for trooping (130,260) in the described step of the profile information of described digital photograph, the particular community that will consider when being logged in the described digital photograph group of described trooping (130,260) creates described situation profile.
8. such as each described method among the claim 1-7, also comprise: in the described step (130 of the described digital photograph group of trooping according to described profile information, 260) before, determine the determinacy of different attribute with respect to different digital photographs, and leach the uncertain information of digital photograph (250) have to(for) particular community.
9. such as each described method among the claim 1-8, wherein according to the troop described step (130 of described digital photograph group of described profile information, 260) comprising: according to one or more attribute definition clusters, and according to the described cluster described digital photograph group of trooping.
10. such as each described method in claim 8 or 9, also comprise: the described step (270) at the representative digital photograph of identifying described cluster based on described profile information from the digital photograph of trooping is added (280) to suitable cluster with described digital photograph that leach, that have uncertain information afterwards.
11. such as each described method among the claim 8-10, wherein determine that with respect to different digital photographs the determinacy of different attribute comprises: determine the entropy of described attribute, wherein high entropy is corresponding to uncertain information.
12. such as each described method among the claim 1-11, wherein identify the described step (140 of representative digital photograph, 270) comprising: identify the photo in each cluster, described photo is best corresponding to the attribute of defined that cluster in described profile information.
13. such as each described method among the claim 1-12, if typing user profiles and subsequently described user profiles being used in according to the troop described step (130 of described digital photograph group of described profile, 260) in, then described method also comprises: receive from the feedback (310) of the representative digital photograph of the user, relevant identification and adjust the weight (320) of each attribute in the described user profiles according to the feedback that receives.
14. the method for claim 1, the described step (100,200) that wherein receives and/or preserve the digital photograph group of identifying comprising: described digital photograph group is uploaded to data storage device.
15. the method for claim 1, wherein receive and/or preserve the described step (100 of the digital photograph group of identifying, 200) comprising: at least one digital photograph uploaded to the data storage device that comprises the digital photograph set of clusters that has existed, and wherein saidly troop (130 according to described profile, 260) described digital photograph group comprises: described at least one digital photograph of uploading is trooped in one of (130,260) cluster in the described digital photograph set of clusters that has existed.
16. an image organizational equipment (500) is suitable for organizing a plurality of digital photographs, described image organizational equipment (500) comprising:
-database (510,511) is suitable for receiving and/or preserving the digital photograph group of identifying,
-receiving element (540) is suitable for receiving the input from the user, and described input comprises the number of clusters that will form and will be used for described digital photograph is trooped to the profile information of the cluster of described quantity from described digital photograph group,
-the unit of trooping (530) is suitable for according to the described profile information described digital photograph group of trooping, and
-recognition unit (520) is suitable for identifying from the digital photograph of trooping based on described profile information the representative digital photograph of described cluster.
17. equipment as claimed in claim 16 (500), wherein said profile information comprises attribute, and described attribute represents the different qualities of digital photograph.
18. equipment as claimed in claim 17 (500), wherein said attribute comprise the characteristic of the outward appearance of the metadata of photo and/or relevant described photo.
19. such as each described equipment (500) among the claim 16-18, wherein weighted associations is arrived some attributes or all properties, thereby defines the importance of each attribute.
20. such as each described equipment (500) among the claim 16-19, the profile information of wherein said reception be definition for the user profiles of the effective community set of specific user, described attribute is defined in the described different qualities of trooping and will consider during described digital photograph group.
21. such as each described equipment (500) among the claim 16-19, the profile information of wherein said reception be definition for the situation profile of the effective community set of particular condition, described attribute is defined in the described different qualities of trooping and will consider during described digital photograph group.
22. equipment as claimed in claim 21 (500), wherein said receiving element (540) is suitable for: troop during the particular community that will consider during described digital photograph group when described receiving element (540) receives described, create described situation profile.
23. such as each described equipment (500) among the claim 16-22, the described unit of trooping (530) also is suitable for: described digital photograph group is trooped in the described unit of trooping (530) according to described profile information before, determine the determinacy of different attribute and leach the photo that has uncertain information for particular community with respect to different digital photographs.
24. such as each described equipment (500) among the claim 16-23, the wherein said unit of trooping (530) also is suitable for: when trooping described digital photograph group according to described profile information, according to one or more attribute definition clusters, and according to the described cluster described digital photograph group of trooping.
25. such as each described equipment (500) in claim 23 or 24, the wherein said unit of trooping (530) also is suitable for: after identifying the representative digital photograph of described cluster based on described profile information from the digital photograph of trooping, add described digital photograph that leach, that have uncertain information to suitable cluster.
26. such as each described equipment (500) among the claim 23-25, wherein saidly determine that with respect to different digital photographs the determinacy of different attribute comprises: determine the entropy of described attribute, wherein high entropy is corresponding to uncertain information.
27. such as each described equipment (500) among the claim 26-25, the representative digital photograph of wherein said identification comprises: identify the photo in each cluster, described photo is best corresponding to the attribute of defined that cluster in described profile information.
28. such as each described equipment (500) among the claim 16-27, if wherein described receiving element (540) receives the user profiles that is used when trooping described digital photograph group according to described profile by the described unit of trooping (530) subsequently, then described receiving element (540) also is suitable for receiving from the feedback user, the relevant representative digital photograph of identifying, and the described unit of trooping (530) also is suitable for adjusting according to the feedback that receives the weight of each attribute in the described user profiles.
29. equipment as claimed in claim 16 (500), wherein said database (510,511) be suitable for preserving the group of the digital photograph of having trooped and receive at least one digital photograph, and the wherein said unit of trooping (530) is suitable for also at least one digital photograph of described reception is trooped in one or more clusters in to the digital photograph of having trooped described group.
30. such as each described equipment (500) among the claim 16-29, wherein said equipment is terminal, such as digital camera, mobile phone or comprise any other terminal of video camera.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/EP2010/057895 WO2011154031A1 (en) | 2010-06-07 | 2010-06-07 | Method and apparatus for organizing images |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102918526A true CN102918526A (en) | 2013-02-06 |
Family
ID=42418352
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010800672988A Pending CN102918526A (en) | 2010-06-07 | 2010-06-07 | Method and apparatus for organizing images |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US20130066872A1 (en) |
| EP (1) | EP2577502A1 (en) |
| CN (1) | CN102918526A (en) |
| WO (1) | WO2011154031A1 (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8891883B2 (en) | 2012-05-15 | 2014-11-18 | Google Inc. | Summarizing a photo album in a social network system |
| KR101522014B1 (en) * | 2013-04-19 | 2015-05-20 | 인텔렉추얼디스커버리 주식회사 | The method, device and recording medium for managing application icon |
| US9208171B1 (en) * | 2013-09-05 | 2015-12-08 | Google Inc. | Geographically locating and posing images in a large-scale image repository and processing framework |
| US10474407B2 (en) | 2013-10-10 | 2019-11-12 | Pushd, Inc. | Digital picture frame with automated interactions with viewer and viewer devices |
| US20150293982A1 (en) * | 2014-04-14 | 2015-10-15 | International Business Machines Corporation | Displaying a representative item for a collection of items |
| WO2015200350A1 (en) | 2014-06-24 | 2015-12-30 | Google Inc. | Ranking and selecting images for display from a set of images |
| US10290019B2 (en) * | 2014-10-24 | 2019-05-14 | Dropbox, Inc. | User re-engagement with online photo management service |
| US11055321B2 (en) * | 2016-11-22 | 2021-07-06 | Photo Butler, Inc. | Rules based albums |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1024437A2 (en) * | 1999-01-26 | 2000-08-02 | Xerox Corporation | Multi-modal information access |
| US20060193538A1 (en) * | 2005-02-28 | 2006-08-31 | Microsoft Corporation | Graphical user interface system and process for navigating a set of images |
| CN101320431A (en) * | 2008-06-13 | 2008-12-10 | 北京航空航天大学 | Self-adapting SAR image classification method based on local standard deviation |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9495583B2 (en) * | 2009-01-05 | 2016-11-15 | Apple Inc. | Organizing images by correlating faces |
-
2010
- 2010-06-07 CN CN2010800672988A patent/CN102918526A/en active Pending
- 2010-06-07 US US13/697,946 patent/US20130066872A1/en not_active Abandoned
- 2010-06-07 EP EP10734710.6A patent/EP2577502A1/en not_active Withdrawn
- 2010-06-07 WO PCT/EP2010/057895 patent/WO2011154031A1/en active Application Filing
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1024437A2 (en) * | 1999-01-26 | 2000-08-02 | Xerox Corporation | Multi-modal information access |
| US20060193538A1 (en) * | 2005-02-28 | 2006-08-31 | Microsoft Corporation | Graphical user interface system and process for navigating a set of images |
| CN101320431A (en) * | 2008-06-13 | 2008-12-10 | 北京航空航天大学 | Self-adapting SAR image classification method based on local standard deviation |
Also Published As
| Publication number | Publication date |
|---|---|
| US20130066872A1 (en) | 2013-03-14 |
| EP2577502A1 (en) | 2013-04-10 |
| WO2011154031A1 (en) | 2011-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102918526A (en) | Method and apparatus for organizing images | |
| US20230319130A1 (en) | Systems and methods for selecting media items | |
| US9223802B2 (en) | Automatic sharing of superlative digital images | |
| US8923570B2 (en) | Automated memory book creation | |
| US8958650B1 (en) | Device and computer readable medium for sharing image content based on collection proximity | |
| US9177206B2 (en) | Automatic photo album creation based on social information | |
| US10459968B2 (en) | Image processing system and image processing method | |
| CN109523237B (en) | Crowd-sourced task pushing method and related device based on user preference | |
| US9247106B2 (en) | Color correction based on multiple images | |
| US9686213B2 (en) | Method and system for account recommendation | |
| US8983150B2 (en) | Photo importance determination | |
| US10977677B2 (en) | Contact importer | |
| US11768871B2 (en) | Systems and methods for contextualizing computer vision generated tags using natural language processing | |
| JP2015141530A (en) | information processing apparatus, score calculation method, program, and system | |
| US10360255B2 (en) | Systems and methods to determine location of media items | |
| US10885619B2 (en) | Context-based imagery selection | |
| US9020863B2 (en) | Information processing device, information processing method, and program | |
| CN104981753B (en) | Method and apparatus for content manipulation | |
| CN107038169B (en) | Object recommendation method and object recommendation device | |
| EP3318021B1 (en) | Managing and indexing communication data, and recommending communication type | |
| KR101749603B1 (en) | Integrated photo management system using photo information and member information, method of photo management and method of social network service using the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130206 |